Reliability of neural activation and connectivity during implicit face emotion processing in youth

Face emotion imaging paradigms are widely used in both healthy and psychiatric populations. Here, in children and adolescents, we evaluate the test-retest reliability of blood oxygenation-level dependent (BOLD) activation and task-based functional connectivity on a widely used implicit face emotion processing task (i.e., gender labeling). Twenty-five healthy youth (M age = 13.97 year s; 60% female) completed two functional magnetic resonance imaging (fMRI) scan sessions approximately two months apart. Participants identified the gender of faces displaying angry, fearful, happy, and neutral emotions. A Bayesian adaptation of the intraclass correlation (ICC) assessed reliability of evoked BOLD activation and amygdala seed-based functional connectivity on task events vs. baseline as well as contrasts between face emotions. For each face emotion vs. baseline, good reliability of activation was demonstrated across key emotion processing regions including middle, medial, and inferior frontal gyri. However, contrasts between face emotions yielded variable results. Contrasts of angry to neutral or happy faces exhibited good reliability of amygdala connectivity to prefrontal regions. Contrasts of fearful to happy faces exhibited good reliability of activation in the anterior cingulate. Findings inform the reproducibility literature and emphasize the need for continued evaluation of task reliability.